🔬 optics

Geometry-Encoded Multireceiver Fluorometry Enables Full-Range Nonlinear Quantification under the Inner Filter Effect

This paper introduces a geometry-encoded multireceiver fluorometry strategy that harnesses the inner filter effect as a quantitative encoding dimension to achieve full-range, nonlinear concentration quantification without requiring sample dilution or separate absorbance measurements.

Original authors: Shuiyi Tan, Nai-Quan Zhu, Olivier J. F. Martin, Yuchao Fu

Published 2026-08-21

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Shuiyi Tan, Nai-Quan Zhu, Olivier J. F. Martin, Yuchao Fu

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the world of light and chemistry, scientists often rely on a simple rule: the brighter a sample glows, the more of a substance it contains. This relationship works beautifully when the substance is sparse, allowing researchers to count molecules by the light they emit. However, this rule breaks down when the sample becomes too crowded. As the concentration of a substance increases, it begins to block its own light. The incoming light gets absorbed before it can reach the deeper parts of the sample, and the light emitted by the molecules gets swallowed up again before it can escape to the detector. This phenomenon, known as the inner filter effect, turns a straightforward measurement into a confusing puzzle. Instead of a steady rise in brightness, the signal peaks and then starts to fall, meaning a very bright reading could actually indicate a low concentration, while a dim reading might signal a dangerously high one. For decades, scientists have tried to fix this by diluting samples or adding extra measurements to correct the math, but these workarounds often complicate the process or disturb the very balance they are trying to measure.

A team of researchers has now proposed a different way to think about this problem. Rather than trying to erase the distortion caused by the inner filter effect, they decided to use it as a feature. Their approach involves looking at the sample not from a single angle, but through multiple "windows" placed at different positions along the path of the light. Imagine a long, clear tube filled with a glowing liquid. If you shine a light into one end, the liquid near the entrance glows brightly, but the light fades as it travels deeper because the liquid absorbs it. If you place a single sensor at the far end, you might miss the glow entirely if the liquid is too thick. But if you place several sensors along the side of the tube, each one sees a different slice of the fading light. The researchers found that by combining the signals from these different positions, they could create a unique fingerprint for every concentration level, even when the total brightness was decreasing.

The team tested this idea using tryptophan, a common building block of proteins that glows under ultraviolet light. They built a custom device with a long viewing cell and positioned multiple sensors along the side to capture the light at different distances from the source. When they measured samples with increasing amounts of tryptophan, a single sensor behaved exactly as expected: the signal rose, hit a peak, and then dropped, creating a situation where one brightness level corresponded to two different concentrations. This ambiguity made it impossible to know the true amount of substance just by looking at one number. However, when they combined the data from two sensors placed at different spots, the ambiguity vanished. The first sensor might see a rising signal while the second saw a falling one, or both might be on different parts of the curve. By feeding these combined readings into a computer model that understood the physics of how light travels through the liquid, the system could pinpoint the exact concentration without any guesswork.

To find the best arrangement for these sensors, the researchers simulated dozens of different setups, moving the virtual windows closer or further apart and changing their sizes. They discovered that the most accurate configuration used one narrow window right at the entrance where the light was strongest and a second, wider window further down the tube. This specific pairing allowed them to measure concentrations ranging from 5 to 125 milligrams per liter with remarkable precision. In their experiments, the system could determine the concentration with an average error of less than 1 milligram per liter, a level of accuracy that held true even when the light was heavily filtered by the sample. They also tested a setup with three sensors, which improved the precision slightly, but they found that adding more than two sensors offered diminishing returns, as the second sensor was the key to solving the initial confusion.

The study demonstrates that by treating the physical layout of the measurement as a variable, scientists can turn a major obstacle into a powerful tool. Instead of fighting the way light fades in a dense sample, this method embraces the fade, using the specific pattern of that fading to calculate the answer. The researchers confirmed that their method works across the entire range of concentrations, including the difficult high-concentration zone where traditional methods fail. They showed that this approach does not require diluting the sample, adding extra chemicals, or performing separate tests to measure how much light is absorbed. While the current work focused on a specific type of protein building block in water, the underlying principle suggests that any fluorescent measurement could potentially be improved by designing the observation geometry to capture the spatial story of the light. This shifts the focus from correcting a broken signal to reading a complex one, offering a new path for analyzing thick, dark, or highly concentrated samples in the future.

Technical Summary: Geometry-Encoded Multireceiver Fluorometry for Nonlinear Quantification

Problem Statement
Fluorescence spectroscopy is widely used for biochemical and clinical applications due to its sensitivity and compact instrumentation. However, the fundamental assumption of a linear relationship between fluorescence intensity and fluorophore concentration holds only in dilute limits. At higher concentrations, the Inner Filter Effect (IFE) distorts this relationship. The primary IFE (pIFE) attenuates the excitation beam before it reaches the observation volume, while the secondary IFE (sIFE) reabsorbs emitted photons. These effects cause the fluorescence response to become sublinear, reach a stationary point (vanishing gradient), and eventually decrease as concentration increases.

This non-monotonic behavior creates a critical metrological failure: a single scalar intensity measurement becomes ambiguous, potentially mapping a high concentration to the same signal as a lower concentration (the "two-valued" problem). Traditional approaches attempt to correct or eliminate IFE through dilution, auxiliary absorbance measurements, or geometric corrections (e.g., front-face detection). However, these methods often introduce workflow complexity, perturb sample equilibria, or remain conditional on specific optical geometries and spectral stability.

Methodology
The authors propose a Geometry-Encoded Multireceiver Fluorometric Strategy that does not attempt to eliminate IFE but instead harnesses the spatial attenuation gradient induced by IFE as an additional quantitative encoding dimension.

Physical Model: The study derives a spatial forward model for lateral (90°) collection. The fluorescence generation line density is modeled as an exponential decay along the excitation axis ( $x$ ), modulated by both excitation attenuation ( $\mu_{ex}$ ) and emission path attenuation ( $\mu_{em}$ ). The detected intensity for a receiver window spanning $[x_1, x_2]$ is the integral of this density, resulting in a response function dependent on the difference of exponentials: $I_f \propto (e^{-\beta c x_1} - e^{-\beta c x_2})$ .
Multireceiver Configuration: Instead of a single detector, the system employs multiple spatially separated receiver windows (e.g., two or three channels) positioned along the excitation axis. Each window integrates a different portion of the spatially varying fluorescence field.
Nonlinear Inversion: Concentration is recovered by solving a nonlinear optimization problem. The system minimizes the residual between a measured intensity vector ( $\mathbf{I}_{meas}$ ) and a predicted vector ( $\mathbf{I}(c)$ ) generated by the forward model. The objective function is $\Delta(c) = \frac{1}{2} \sum |I_j(c) - I_{meas, j}|^2$ .
Optimization and Solvers: The study evaluates three nonlinear solvers (Newton, gradient-descent, and Levenberg-Marquardt) to find the global minimum. It also employs a discrete screening of 35 two-receiver geometries and a continuous genetic algorithm to identify configurations that minimize uncertainty and maximize sensitivity.
Metrics: Performance is evaluated using profile-likelihood error half-width ( $E_{95}$ ) to capture asymmetric uncertainty and noise-normalized sensitivity ( $S_N$ ) to assess local derivative-to-noise ratios.

Key Contributions

Paradigm Shift: The paper reframes IFE from a source of error to a source of information. By treating spatial attenuation as a concentration-dependent coordinate, the method achieves global identifiability where single-channel methods fail.
Restoration of Uniqueness: Theoretical analysis and simulations demonstrate that while a single receiver channel produces a non-monotonic, two-valued response curve, the joint vector of two spatially separated receivers collapses the solution space to a unique global minimum, eliminating concentration ambiguity across the full range.
Geometry Optimization: The study identifies that receiver geometry is a first-order analytical design variable. It distinguishes between configurations that minimize global uncertainty (e.g., a narrow entrance window paired with a downstream window) and those that maximize local sensitivity.
Algorithmic Robustness: The work confirms that once the geometric ambiguity is resolved by multireceiver design, standard nonlinear solvers converge reliably to the true concentration regardless of initialization (low or high concentration).

Experimental Results
The strategy was validated using L-Tryptophan as a model fluorophore in aqueous solution, spanning a concentration range of 5 to 125 mg/L.

Single vs. Multichannel: Single-window calibration exhibited vanishing-gradient boundaries and two-valued inversions. In contrast, a two-channel configuration restored global uniqueness.
Geometric Screening: Among 35 tested two-receiver geometries, the configuration with windows at [0, 1] mm and [4, 7] mm achieved the minimum average uncertainty ( $E_{95} = 0.884$ mg/L). The configuration at [0, 3] mm and [4, 7] mm achieved the maximum sensitivity ( $S_N = 3.832$ a.u./(mg/L)).
Full-Range Quantification: Using 125 experimental measurements, the multireceiver inversion achieved an $R^2$ of 0.997. The Root Mean Square Error (RMSE) was reduced to 1.934 mg/L, and the Mean Absolute Error (MAE) to 1.440 mg/L.
Diminishing Returns: Adding a third receiver further improved precision (reducing RMSE by 17.1% and MAE by 20.2%), but the primary leap in performance was achieved by moving from one to two channels, which established global uniqueness.

Significance and Claims
The paper claims to provide a general metrology framework for fluorescence analysis under strong IFE conditions without requiring sample dilution, separate absorbance measurements, or piecewise calibration. By converting spatial attenuation into a multidimensional concentration coordinate, the approach extends the quantitative working range of fluorometry.

The authors emphasize that this method fundamentally differs from post-hoc correction techniques; it retains the spatial attenuation as intrinsic information and estimates concentration directly from a nonlinear spatial forward model. While the current validation is limited to aqueous tryptophan, the authors assert that the optical inverse-problem principle is established, though they note that future work must address matrix effects (scattering, binding, spectral shifts) in complex biological samples like blood or serum before clinical translation. The immediate next steps identified are rigorous matrix validation, joint treatment of scattering and sIFE, and hardware implementation with fixed multielement detectors.

Technical Summary: Geometry-Encoded Multireceiver Fluorometry for Nonlinear Quantification

More like this